1376 Letters to the Editor
events beyond the acute diverticulitis attack and outside the digestive system. LISA L. STRATE, MD, MPH Department of Medicine University of Washington School of Medicine Seattle, Washington RUNE ERICHSEN, MD, PhD Department of Clinical Epidemiology Aarhus University Hospital Aarhus, Denmark JOHN A. BARON, MD, MS, MSc Department of Medicine University of North Carolina Chapel Hill, North Carolina HENRIK TOFT-SØRENSEN, MD, PhD, DMS Department of Clinical Epidemiology Aarhus University Hospital Aarhus, Denmark
References 1.
Strate LL, et al. Clin Gastroenterol Hepatol 2014;12:1695–1701.e1.
2.
Strate LL, et al. Am J Gastroenterol 2012;107:1486–1493.
3.
Fumery M, et al. J Crohns Colitis 2014;8:469–479.
4.
Kappelman MD, et al. Gut 2011;60:937–943.
Conflicts of interest The authors disclose no conflicts. http://dx.doi.org/10.1016/j.cgh.2015.04.007
Continuous-flow Left Ventricular Assist Device, Valvular Disease, and Acquired von Willebrand Syndrome Dear Editor: We read the recent article by Shrode et al,1 which intended to characterize gastrointestinal (GI) bleeds and thromboembolic events that occurred in patients with continuous-flow left ventricular assist devices (CF-LVADs) and compare them with patients receiving anticoagulation after cardiac valve replacement. We commend the authors for their completion of a wellexecuted evaluation of a difficult and timely topic. We further discuss the propensity to GI bleeding related to acquired von Willebrand syndrome (AcVWS) in patients with CF-LVADs. AcVWS refers to defects in von Willebrand factor (vWF) concentration, structure, or function that are not inherited but are consequences of other medical disorders. In patients with CF-LVADs and valvular disease, the increased shear stress can increase the proteolysis of vWF by ADAMTS13 enough to deplete large vWF multimers (HMW) and thereby produce a bleeding diathesis that resembles type 2A von Willebrand disease. The multimers with higher molecular weight can capture more platelets, like larger cobwebs, and are more effective in adhesion and aggregation. Therefore, loss of HMW
Clinical Gastroenterology and Hepatology Vol. 13, No. 7
multimers impairs the early steps of primary hemostasis and causes bleeding diathesis consequently.2 Multiple studies reported that AcVWS, diagnosed by the decrease of HMW vWF multimers, develops in all patients on CF-LVAD support.3,4 This hemostatic abnormality may explain the excessive bleeding in the CF-LVAD patients. Bleeding complications associated with these devices range from as low as 17.5% to as high as 63 events/100 patient-years.3 Despite the uniform nature of these findings, not every patient on CF-LVAD develops major bleeding complications; therefore, HMW vWF multimer deficiency alone does not appear to fully explain why some CF-LVAD recipients bleed when others do not. The presence of angiodysplasia may determine whether a patient has bleeding after CF-LVAD placement. Bleeding in the setting of HMW vWF multimer loss may only occur if the patient has coexisting GI angiodysplasia that is dependent on HMW vWF multimers to maintain hemostasis.4 Immediate recovery of AcVWS after CF-LVAD removal or heart transplant has been shown to decrease the rates of bleeding.5 In 1958, Heyde et al6 reported a high incidence of GI bleeding in aortic stenosis patients (Heyde syndrome). In the 1990s, Warkentin et al7 raised the question of whether stenotic aortic valves predisposed patients to the development of AcVW syndrome, which could contribute to high bleeding from GI angiodysplasia. In multiple case series, correction of valvular disease appears to result in normalization of vWF levels and vWF multimer composition, likely through resolution of induced increases in mechanical stress. Deficiency of the vWF multimers undergoes long-term normalization after valve replacement.8 In the current study by Shrode et al,1 we consider that the higher incidence of GI bleeding in the CF-LVAD group as compared with the control group was due to the presence of AcVWS in the former as compared with their counterparts (control group), where resolution of this syndrome has been documented. Because AcVWS is common in patients undergoing placement of CF-LVADs and risk stratification with assessment of the GI tract with upper endoscopy, wireless capsule endoscopy and colonoscopy looking for possible source of bleeding such as angiodysplasia, polyps, or masses could be a novel approach to decrease the morbidity and high incidence of GI bleeding in this population. FRANK CZUL, MD JAMIE S. BARKIN, MD Division of Gastroenterology Department of Medicine University of Miami Miller School of Medicine Miami, Florida
References 1.
Shrode CW, et 12:1461–1467.
al.
Clin
Gastroenterol
Hepatol
2.
Nichols WL, et al. Haemophilia 2008;14:171–232.
2014;
July 2015 3.
Uriel N, et al. J Am Coll Cardiol 2010;56:1207–1213.
4.
Crow S, et al. Ann Thorac Surg 2010;90:1263–1269.
5.
Davis ME, et al. ASAIO J 2015;61:e1–4.
6.
Heyde E. N Engl J Med 1958;259:196–200.
7.
Warkentin TE, et al. Lancet 1992;340:35–37.
8.
Scheffer SM, et al. Ann Thorac Surg 1986;42:477–480.
Letters to the Editor 1377
Conflicts of interest The authors disclose no conflicts. http://dx.doi.org/10.1016/j.cgh.2014.11.023
Measuring the Value of Colonoscopists’ Performance Dear Editor: The article by Gohel at al1 not only addresses the relevant relationship between polypectomy rate (PR) and adenoma detection rate (ADR) but also provides important clinical insights into the value of colonoscopy performance among a large cohort of endoscopists. With the growing emphasis on value-based payment, the value of clinicians’ performance is now subject to greater scrutiny. In healthcare, value is defined as quality of care adjusted for cost.2,3 In the context of colonoscopy practice, one approach to expressing value is the ratio of adenoma detection, a measure of quality, to polyp detection, which can serve as a surrogate estimate of procedure cost. The principal cost components of a colonoscopy are technical facility charge, endoscopist professional charge, pathology charge, and anesthesia charge.4 In endoscopy units in which sedation is administered by the endoscopist, thereby avoiding any additional anesthesiologist professional fee, the number of histology specimen containers used, which generally correlates closely with PR, represents the major source of variable costs per procedure. Therefore, with these assumptions in mind, the ratio of ADR to PR can be used to estimate the value of a colonoscopist’s performance, ie, quality/cost. In the article by Gohel et al,1 the overall mean ADR/PR ratio for the 65 colonoscopists studied was 58%, with a broad variation from 0% to 97% and standard deviation of 20%. The coefficient of variation (ratio of the standard deviation to the mean) was 34%; by convention, a coefficient of variation >15% represents broad spread, thereby indicating a high level of variation in value of endoscopists’ performances. Figure 1 illustrates the relationship between ADR and PR for the author’s study group. The horizontal line separates endoscopists with ADR above and below the mean (25%). As shown, the cohort of endoscopists who achieve an acceptable level of quality (ADR >25%) can be further classified as higher value or lower value performers on the basis of whether they have high PR (thereby incurring high pathology charges) or low PR (incurring lesser pathology charges). Although there are many measures of colonoscopy quality other than ADR and many contributors to cost other than pathology charges, the data provided by Gohel
Figure 1. Relationship between ADR and PR for colonoscopists. The horizontal line classifies endoscopists with ADR above and below the mean (25%); the vertical line classifies those endoscopists with ADR >25% with lower and higher PRs, ie, lower value and higher value.
et al1 illustrate one approach to numerically quantifying the value of colonoscopy performance. With the introduction of the Affordable Care Act and the shift toward value-based reimbursement in place of a fee-for-service model of payment, value measurement, not just in endoscopy but in all aspects of clinical care, will assume increasing importance. GAVIN C. HAREWOOD, MD, MBA, MSc, FASGE Department of Gastroenterology and Hepatology Beaumont Hospital Dublin, Ireland
References 1. 2.
Gohel TD, et al. Clin Gastroenterol Hepatol 2014;12:1137–1142. Porter ME. N Engl J Med 2010;363:2477–2481.
3.
Dulai PS, et al. Clin Gastroenterol Hepatol 2012;10:609–611.
4.
Allen JI. Gastroenterology 2014;146:573–575.
Conflicts of interest The author discloses no conflicts. http://dx.doi.org/10.1016/j.cgh.2014.07.048
Reply. We would like to thank Dr Harewood1 for his interest in our study2 and interpretation of our results. He raises an important point in this era of cost-conscious health care delivery. He makes an interesting assumption that ratio of adenoma detection rate (ADR) to polypectomy rate (PR) can be used to estimate the value of a colonoscopists’ performance ie, higher value performers have high ADR and lower PR thereby incurring lower pathology charges as opposed to lower value performers who have low ADR and high PR. In other words, high value performers are better at differentiating neoplastic versus non-neoplastic polyps on inspection and they preferentially remove neoplastic appearing polyps. There is considerable interest in this area as evidenced by the resect and discard approach to treat diminutive polyps. In this method, the histology of a diminutive polyp is assessed during endoscopy, a highresolution photograph is taken, and the polyp is then
events beyond the acute diverticulitis attack and outside the digestive system. LISA L. STRATE, MD, MPH Department of Medicine University of Washington School of Medicine Seattle, Washington RUNE ERICHSEN, MD, PhD Department of Clinical Epidemiology Aarhus University Hospital Aarhus, Denmark JOHN A. BARON, MD, MS, MSc Department of Medicine University of North Carolina Chapel Hill, North Carolina HENRIK TOFT-SØRENSEN, MD, PhD, DMS Department of Clinical Epidemiology Aarhus University Hospital Aarhus, Denmark
References 1.
Strate LL, et al. Clin Gastroenterol Hepatol 2014;12:1695–1701.e1.
2.
Strate LL, et al. Am J Gastroenterol 2012;107:1486–1493.
3.
Fumery M, et al. J Crohns Colitis 2014;8:469–479.
4.
Kappelman MD, et al. Gut 2011;60:937–943.
Conflicts of interest The authors disclose no conflicts. http://dx.doi.org/10.1016/j.cgh.2015.04.007
Continuous-flow Left Ventricular Assist Device, Valvular Disease, and Acquired von Willebrand Syndrome Dear Editor: We read the recent article by Shrode et al,1 which intended to characterize gastrointestinal (GI) bleeds and thromboembolic events that occurred in patients with continuous-flow left ventricular assist devices (CF-LVADs) and compare them with patients receiving anticoagulation after cardiac valve replacement. We commend the authors for their completion of a wellexecuted evaluation of a difficult and timely topic. We further discuss the propensity to GI bleeding related to acquired von Willebrand syndrome (AcVWS) in patients with CF-LVADs. AcVWS refers to defects in von Willebrand factor (vWF) concentration, structure, or function that are not inherited but are consequences of other medical disorders. In patients with CF-LVADs and valvular disease, the increased shear stress can increase the proteolysis of vWF by ADAMTS13 enough to deplete large vWF multimers (HMW) and thereby produce a bleeding diathesis that resembles type 2A von Willebrand disease. The multimers with higher molecular weight can capture more platelets, like larger cobwebs, and are more effective in adhesion and aggregation. Therefore, loss of HMW
Clinical Gastroenterology and Hepatology Vol. 13, No. 7
multimers impairs the early steps of primary hemostasis and causes bleeding diathesis consequently.2 Multiple studies reported that AcVWS, diagnosed by the decrease of HMW vWF multimers, develops in all patients on CF-LVAD support.3,4 This hemostatic abnormality may explain the excessive bleeding in the CF-LVAD patients. Bleeding complications associated with these devices range from as low as 17.5% to as high as 63 events/100 patient-years.3 Despite the uniform nature of these findings, not every patient on CF-LVAD develops major bleeding complications; therefore, HMW vWF multimer deficiency alone does not appear to fully explain why some CF-LVAD recipients bleed when others do not. The presence of angiodysplasia may determine whether a patient has bleeding after CF-LVAD placement. Bleeding in the setting of HMW vWF multimer loss may only occur if the patient has coexisting GI angiodysplasia that is dependent on HMW vWF multimers to maintain hemostasis.4 Immediate recovery of AcVWS after CF-LVAD removal or heart transplant has been shown to decrease the rates of bleeding.5 In 1958, Heyde et al6 reported a high incidence of GI bleeding in aortic stenosis patients (Heyde syndrome). In the 1990s, Warkentin et al7 raised the question of whether stenotic aortic valves predisposed patients to the development of AcVW syndrome, which could contribute to high bleeding from GI angiodysplasia. In multiple case series, correction of valvular disease appears to result in normalization of vWF levels and vWF multimer composition, likely through resolution of induced increases in mechanical stress. Deficiency of the vWF multimers undergoes long-term normalization after valve replacement.8 In the current study by Shrode et al,1 we consider that the higher incidence of GI bleeding in the CF-LVAD group as compared with the control group was due to the presence of AcVWS in the former as compared with their counterparts (control group), where resolution of this syndrome has been documented. Because AcVWS is common in patients undergoing placement of CF-LVADs and risk stratification with assessment of the GI tract with upper endoscopy, wireless capsule endoscopy and colonoscopy looking for possible source of bleeding such as angiodysplasia, polyps, or masses could be a novel approach to decrease the morbidity and high incidence of GI bleeding in this population. FRANK CZUL, MD JAMIE S. BARKIN, MD Division of Gastroenterology Department of Medicine University of Miami Miller School of Medicine Miami, Florida
References 1.
Shrode CW, et 12:1461–1467.
al.
Clin
Gastroenterol
Hepatol
2.
Nichols WL, et al. Haemophilia 2008;14:171–232.
2014;
July 2015 3.
Uriel N, et al. J Am Coll Cardiol 2010;56:1207–1213.
4.
Crow S, et al. Ann Thorac Surg 2010;90:1263–1269.
5.
Davis ME, et al. ASAIO J 2015;61:e1–4.
6.
Heyde E. N Engl J Med 1958;259:196–200.
7.
Warkentin TE, et al. Lancet 1992;340:35–37.
8.
Scheffer SM, et al. Ann Thorac Surg 1986;42:477–480.
Letters to the Editor 1377
Conflicts of interest The authors disclose no conflicts. http://dx.doi.org/10.1016/j.cgh.2014.11.023
Measuring the Value of Colonoscopists’ Performance Dear Editor: The article by Gohel at al1 not only addresses the relevant relationship between polypectomy rate (PR) and adenoma detection rate (ADR) but also provides important clinical insights into the value of colonoscopy performance among a large cohort of endoscopists. With the growing emphasis on value-based payment, the value of clinicians’ performance is now subject to greater scrutiny. In healthcare, value is defined as quality of care adjusted for cost.2,3 In the context of colonoscopy practice, one approach to expressing value is the ratio of adenoma detection, a measure of quality, to polyp detection, which can serve as a surrogate estimate of procedure cost. The principal cost components of a colonoscopy are technical facility charge, endoscopist professional charge, pathology charge, and anesthesia charge.4 In endoscopy units in which sedation is administered by the endoscopist, thereby avoiding any additional anesthesiologist professional fee, the number of histology specimen containers used, which generally correlates closely with PR, represents the major source of variable costs per procedure. Therefore, with these assumptions in mind, the ratio of ADR to PR can be used to estimate the value of a colonoscopist’s performance, ie, quality/cost. In the article by Gohel et al,1 the overall mean ADR/PR ratio for the 65 colonoscopists studied was 58%, with a broad variation from 0% to 97% and standard deviation of 20%. The coefficient of variation (ratio of the standard deviation to the mean) was 34%; by convention, a coefficient of variation >15% represents broad spread, thereby indicating a high level of variation in value of endoscopists’ performances. Figure 1 illustrates the relationship between ADR and PR for the author’s study group. The horizontal line separates endoscopists with ADR above and below the mean (25%). As shown, the cohort of endoscopists who achieve an acceptable level of quality (ADR >25%) can be further classified as higher value or lower value performers on the basis of whether they have high PR (thereby incurring high pathology charges) or low PR (incurring lesser pathology charges). Although there are many measures of colonoscopy quality other than ADR and many contributors to cost other than pathology charges, the data provided by Gohel
Figure 1. Relationship between ADR and PR for colonoscopists. The horizontal line classifies endoscopists with ADR above and below the mean (25%); the vertical line classifies those endoscopists with ADR >25% with lower and higher PRs, ie, lower value and higher value.
et al1 illustrate one approach to numerically quantifying the value of colonoscopy performance. With the introduction of the Affordable Care Act and the shift toward value-based reimbursement in place of a fee-for-service model of payment, value measurement, not just in endoscopy but in all aspects of clinical care, will assume increasing importance. GAVIN C. HAREWOOD, MD, MBA, MSc, FASGE Department of Gastroenterology and Hepatology Beaumont Hospital Dublin, Ireland
References 1. 2.
Gohel TD, et al. Clin Gastroenterol Hepatol 2014;12:1137–1142. Porter ME. N Engl J Med 2010;363:2477–2481.
3.
Dulai PS, et al. Clin Gastroenterol Hepatol 2012;10:609–611.
4.
Allen JI. Gastroenterology 2014;146:573–575.
Conflicts of interest The author discloses no conflicts. http://dx.doi.org/10.1016/j.cgh.2014.07.048
Reply. We would like to thank Dr Harewood1 for his interest in our study2 and interpretation of our results. He raises an important point in this era of cost-conscious health care delivery. He makes an interesting assumption that ratio of adenoma detection rate (ADR) to polypectomy rate (PR) can be used to estimate the value of a colonoscopists’ performance ie, higher value performers have high ADR and lower PR thereby incurring lower pathology charges as opposed to lower value performers who have low ADR and high PR. In other words, high value performers are better at differentiating neoplastic versus non-neoplastic polyps on inspection and they preferentially remove neoplastic appearing polyps. There is considerable interest in this area as evidenced by the resect and discard approach to treat diminutive polyps. In this method, the histology of a diminutive polyp is assessed during endoscopy, a highresolution photograph is taken, and the polyp is then
